U-Th-Pb AND Rb-Sr SYSTEMATICS OF ALLENDE FUN CAI CMS-1

46th Lunar and Planetary Science Conference (2015)
2355.pdf
U-Th-Pb AND Rb-Sr SYSTEMATICS OF ALLENDE FUN CAI CMS-1. Y. Amelin1, C. D. Williams2,3 and M.
Wadhwa2, 1Research School of Earth Sciences, Australian National University, Canberra, Australia
([email protected]), 2School of Earth and Space Exploration, Arizona State University, Tempe, AZ 852871404, USA, 3Department of Earth and Planetary Sciences, University of California at Davis, CA 95616, USA.
Introduction: Among the calcium-aluminum-rich
inclusions (CAIs) there exists a small subgroup referred to as FUN CAIs (for Fractionation and Unidentified Nuclear effects [1]) that exhibit unusually large
mass-dependent fractionations and mass-independent
isotopic variations [1-3]. Understanding the origin,
conditions of formation, genetic relationship of these
inclusions with other CAIs and the time of formation
has been hampered by the rarity of these objects: since
the first discovery of two FUN CAIs 37 years ago [1],
only about 20 such CAIs have been identified [3-4]..
Variable and mostly low compared to “normal”
CAIs 26Al/27Al in FUN CAIs derived from their internal 26Al-26Mg isochrons [4, 5] suggests that these CAIs
formed either prior to, or simultaneously with, injection and homogenization of freshly synthesized 26Al
into the protosolar nebula [6]. Alternatively, the FUN
CAIs could have formed significantly later, or experienced widespread resetting of the 26Al-26Mg isotopic
system. A recent investigation of the 182Hf-182W relative chronometer in a single FUN CAI pointed to its
formation contemporaneously with other CAIs in the
nebula, and additionally suggested heterogeneous distribution of 26Al [7]. The uncertainty in the formation
age of FUN CAIs could be resolved by dating with the
U-Pb isotopic chronometer, but such determinations
have been precluded by very low concentrations of U
[7, 8]. Furthermore, only a few of the identified FUN
CAIs are large enough to be analyzed for multiple geochemical and isotopic systems, especially isotopic dating using macroscopic (more precise but destructive)
methods. Here we report preliminary U-Th-Pb and RbSr data for a recently discovered FUN CAI (CMS-1)
discovered in the Allende meteorite.
Sample: Allende FUN CAI CMS-1 is an irregularly shaped compact Type A inclusion. Its primary mineralogy consists of Ti-Al-rich pyroxene, melilite,
abundant inclusions of spinel [4, 9]. Previous studies
have reported on its Ti, Si, Mg, O isotopic compositions [4,9-10] as well as REE microdistributions [11].
The bulk REE pattern (based on estimated modal
abundances of phases) is relatively flat with a small
negative Ce anomaly and is enriched ~20× relative to
CI chondrites [11].
Methods: A ~5 mg piece was gently crushed in a
mortar and pestle. Coarse (>100 µm, 2.33 mg) and fine
(2.00 mg) fractions were separated for U-Pb isotope
analyses. These fractions were sequentially leached in
cold dilute HNO3 (W1), hot 6M HCl (W2) and hot
dilute HF (W3), leaving the residue (R) that was dissolved in hot concentrated HF. The leachates and the
residues were spiked with 202Pb-205Pb-229Th-233U-236U,
and Pb, U and Th were separated and analyzed using
procedures similar to [12, 13]. The bulks after U, Th
separation were spiked and further analysed for Rb-Sr
isotopes using the procedures of [14].
Results: Uranium concentrations were calculated
by numeric recombination of the leachates and residues. The recombined U concentration in the coarse
fraction is 1.2 ppb, while the fine fraction contains 1.8
ppb; these concentrations are 30-100 times lower than
in “normal” CAIs. A large portion (69-76%) of the U
is located within the easily soluble fraction, notably
W1. All other fractions contain less than 0.2 ppb U.
Since easily leachable U is likely to be partially transferred from the matrix and/or introduced by handling,
1.2 ppb is considered an upper limit for the U concentration in this CAI.
The total (recombined) concentration of Th is 440620 ppb, consistent with the ~20x enrichment (relative
to CI) in REE [11]. Th is distributed uniformly
throughout all leaching steps. Assuming that the parent
material of this CAI had a CI-like Th/U ratio (~4), its
original U concentration is expected to be 100-150 ppb
U, typical for “normal” CAIs.
Lead isotope ratios associated with the 238U-206Pb
and 235U-207Pb decay schemes (Fig.1) are rather un-
Figure 1. 207Pb/206Pb vs. 204Pb/206Pb isochron for the washes
and residues of the CAI CMS-1.
radiogenic (206Pb/204Pb between 10.6 and 16.8 for the
leachates, and close to 15 for the residues) and are consistent with mixing between radiogenic Pb produced
46th Lunar and Planetary Science Conference (2015)
by the decay of the small amount of U present in the
sample and primordial Pb, or Pb transferred from the
matrix into the inclusion. There is no evidence for any
modern terrestrial contamination. The y-intercept of
the 207Pb/206Pb vs. 204Pb/206Pb linear array (if interpreted as an isochron) corresponds to an age of 4586±79
Ma when calculated using 238U/235U=137.79 (the true
238
U/235U could not be determined due to very small
quantity of U available). This age is consistent with the
age of “normal” CAIs, but is too imprecise to provide a
useful constraint on the time of formation for CAI
CMS-1, or FUN CAIs in general.
In contrast to the radiogenic ratios in the U-Pb system, the isotope ratios associated with the 232Th-208Pb
system are much more radiogenic (208Pb/204Pb between
53 and 451 in the leachates and 734-788 in the residues). The possibility of using this system for age determination is, however, comptomised by possible
fractionation between Th and radiogenic Pb, both natural and induced by leaching.
Figure 2. 87Sr/86Sr vs. 87Rb/86Sr isochron for the leachates
and residues of the CAI CMS-1. Reference line corresponds
to the radiogenic growth of 87Sr over 4567 Ma.
The Rb-Sr data for three leachates and residue of
the coarse fraction show some scatter in the isochron
diagram (Fig. 2) and yield apparent initial 87Sr/86Sr
between 0.69839 and 0.69900. The scatter is possibly
caused by Rb-Sr fractionation during leaching, in particular in the last (HF) leaching step. The re-combined
initial 87Sr/86Sr of 0.69884 agrees with the lowest initial 87Sr/86Sr reported for “normal” CAIs [15].
The 84Sr/86Sr data, normalized to 86Sr/88Sr=0.1194,
are shown in Fig. 3 in comparison with values for
leachates (Wash 1 and 2) and residues for “normal”
CAIs from the chondrite NWA 4502 [14]. The leachates and residues of CMS-1 show a consistent deficit of
84
Sr of 5ε (relative to the SRM987 standard), which is
identical to the 84Sr deficit reported for the Allende
FUN CAI C1 [16]. Whether this apparent deficit is real
or is caused by an anomaly in 88Sr (which is one of the
normalizing isotopes for the fractionation correction) is
being investigated. The uniformity of the 84Sr/86Sr ratio
between the fractions is important because it indicates
2355.pdf
Figure 3. 84Sr/86Sr ratios in the Allende CAI CMS-1 (this
study) and CAIs from the CV chondrite NWA 4502 [14].
limited extent, or absence, of Sr isotope exchange between the CAI and the matrix.
Discussion: One of the most intriguing observations in this study is the evidence for extreme elemental fractionation between REE and Th, on one
hand, and U, on the other. Considering “flat” distribution pattern of REE [9], this fractionation is unlikely to
be caused by evaporation and/or condensation processes and the difference in volatility. Nucleosynthetic
origin of low U abundance is also unlikely, since U
and Th are thought to be produced in the same rprocess, and the ratio of their yields have little dependence on the conditions, e.g. neutron flux [17, 18]. In
terrestrial surface environments, U is known to very
efficiently fractionate from REE, Th and other elements by leaching under high oxygen fugacity [e.g.
19]. Similar interaction with an aqueous fluid in the
nebular environment, either in the protosolar nebula or
near the place of primary condensation, could have
caused U removal from the parent material of the FUN
CAIs. Although the reason for the measured depletion
in U is not clear at present, it is noted that the ~5 mg
fraction of CMS-1 analyzed thus far may not be representative of the bulk inclusion. Further analyses will be
required to address this issue.
References: [1] Wasserburg G.J. et al. (1977) GRL 4,
299-302. [2] Papanastassiou D.A. & Brigham C.A. (1989)
ApJ 338, L37–L40. [3] Krot A.N. et al. (2014) GCA 145,
206–247. [4] Williams C.D. et al. (2012) 75th MetSoc Abstract #5102. [5] Park C. et al. (2013) 76th MetSoc Abstract
#5085. [6] Sahijpal S. & Goswami J.N. (1998) ApJ 509,
L137–L140. [7] Holst J.C. et al. (2013) PNAS 110, no. 22,
8819–8823. [8] Chen J.H. & Wasserburg G.J. (1981) EPSL
52, 1-15. [9] Williams C.D. et al. (2013) LPS 44, Abstract #
2435. [10] Williams C.D. et al. (2014) LPS 45, Abstract
#2146. [11] Williams C.D. et al. (2013) 76th MetSoc, Abstract #5108. [12] Amelin Y. et al. (2010) EPSL 300, 343350. [13] Amelin Y. et al. (2013) LPS 44, Abstract #2690.
[14] Amelin Y. et al. (2014) 77th MetSoc, Abstract # 5244.
[15] Gray C.M. et al. (1973) Icarus 20, 213-239. [16]
Papanastassiou D.A. & Wasserburg G.J. (1978) GRL 5, 595598. [17] Goriely S. and Arnould M. (2001) AA 379, 11131122. [18] Lingenfelter R.E. et al. (2003) ApJ 591, 228-237.
[19] Neymark L.A. & Amelin Y. (2008) GCA 72, 2067–
2089.